INDEX

Explanations

strong positive opinions or evaluations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating important social commentary or issues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

usercontent

-0.71

rug

-0.60

upt

-0.59

dule

-0.57

olitan

-0.56

Edit

-0.55

 reads

-0.55

('

-0.55

regate

-0.54

 preparation

-0.54

POSITIVE LOGITS

 borne

0.97

 echoed

0.93

 reflected

0.91

 embodied

0.80

 enshr

0.80

 contrasted

0.79

 bolstered

0.78

 manifested

0.78

 compounded

0.77

ģĸ

0.77

Activations Density 0.285%

strong positive opinions or evaluations

phrases indicating important social commentary or issues

No Comments

No Known Activations